GRANTING ACCESS: LEVERAGING SOCIAL PAYMENTS TO EXPAND DIGITAL FINANCIAL INCLUSION IN COTE D’IVOIRE

Field Note 11

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

2

CONTEXT

The role of cash transfers for poverty reduction has expanded globally over the past two decades. The latest State of Social Safety Nets survey reports that as of 2015, nearly every country around the world, including many in Sub-Saharan Africa, has at least one social safety net program in place. Among these programs, there is a marked trend away from providing food and other in-kind support to delivering cash transfers to the poorest households. Overall, these programs support nearly 2 billion individuals in developing countries. While these cash transfers are expanding, the majority still reach only one tenth of the poorest 20 percent of the population, in part due to a lack of mechanisms to target and deliver the benefits.1 However, the increasing availability of mobile phones and mobile-based digital payment channels offer new ways to improve the delivery of cash transfers while also supporting the financial inclusion of the poorest households living in remote areas. For example, in Colombia, the DaviPlata mobile platform of Banco Davivienda was used to pay almost 1 million beneficiaries of the Familias in Accion program. In Uganda, the Social Assistance Grants for Empowerment (SAGE) program was successfully piloted using a SIM card-based cash payment system building on MTN’s mobile payments infrastructure, delivering bimonthly payments to over 560,000 beneficiaries.2

Until very recently, such unconditional cash transfer (UCT) programs have been all but nonexistent in the West African Economic and Monetary Union (WAEMU).3 This started changing in 2012 when the World Bank together with the concerned ministries initiated UCT programs in Senegal, Guinea-Bissau, and Côte d’Ivoire. The initial programs were small-scale and designed as hand-to-hand cash payments, and quickly brought forth concerns over misappropriation.

They proved difficult to monitor effectively, were expensive, and did not lead to financial inclusion. Recipients often found it hard to manage their benefits and to save, and were more exposed to theft, among other issues.4

While the World Bank expanded its lending operations to social safety nets in West Africa, in parallel, IFC, together with the Mastercard Foundation, started supporting the use of mobile money wallets in WAEMU markets in 2013 through market-level activities and proof-of-concept pilots with the private and public sectors. While less mature than established digital finance markets such as Kenya or Tanzania, WAEMU markets are nascent but rapidly evolving markets in terms of digital financial services and new payment channel development5. Côte d’Ivoire, for instance, has nearly 12 million registered mobile money users (between Orange Money, MTN Mobile Money and Etisalat’s Moov Money) as of March 2018 (ARTCI data), compared to less than 3 million people with bank accounts.6

Côte d’Ivoire is the regional leader and the hub for digital financial services development in WAEMU at both public and private level. Since 2011 the Ministry of Education for instance has been involved in mobile money school fee payments. For the year 2015-2016, 99.3% of Côte d’Ivoire’s 1.7 million secondary school students paid their annual school registration fee via mobile money7. In the first quarter of 2018, MTN launched the first mobile savings and credit service in the region, in partnership with a local bank. Financial institutions are also playing an increasing role, with the microfinance institution Baobab and the bank Société Generale having launched their agency banking models.

1 World Bank (2015) The State of Social Safety Nets 2015. Washington DC: World Bank.2 Merttens, Fred et al (2016). Evaluation of the Uganda Social Assistance Grants For Empowerment (Sage) Programme. Endline programme

operations performance. Oxford Policy Management.3 WAEMU countries include Togo, Benin, Côte d’Ivoire, Senegal, Niger, Mali, Burkina Faso, and Guinea-Bissau4 Pickens, Mark, David Porteous, and Sarah Rotman (2009). “Banking the Poor via G2P Payments.” Washington DC: World Bank. 5 An e-money account/wallet that is primarily accessed using a mobile phone that is held with the e-money issuer.6 World Bank (2018) Global Findex and BCEAO (2018) Etat Des Services Financiers Par Téléphonie Mobile dans L’UEMOA en 20177 GSMA data. https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2015/10/2015_GSMA_Paying-school-fees-with-mobile-

money-in-Cote-dIvoire.pdf

3

8 Republic of Côte d’Ivoire (2013) “Stratégie Nationale de Protection Sociale.9 World Bank (2015). Republic of Côte d’Ivoire Project Appraisal Document, Productive Social Safety Net Project. Report No: PAD1189.

Washington DC: World Bank. The International Development Association (IDA) is the part of the World Bank that helps the world’s poorest countries.

10 Previous experiences in Cote d’Ivoire included small pilots such as MTN and the World Food Program during the last civil unrest.

As part of the Government of Côte d’Ivoire’s National Social Protection Strategy,8 a new targeted social safety net program using cash transfers was launched in 2015, supported by IDA (International Development Association) financing9. In March 2017, the first 5,000 beneficiary households (30,000 individual beneficiaries) received a quarterly allowance of 36,000 FCFA (approx. USD 60) in their Orange Money electronic wallets which was scaled up to 35,000 beneficiary households (or 210,000 individual beneficiaries). This was the first time that such a large number of Ivorian households received cash transfers through mobile money technology 10.

This field note summarizes the steps taken by the World Bank and IFC to operationalize the mobile money-based government-to-person (G2P) transfers together with the government of Côte d’Ivoire. While there were many steps in the course towards implementation, this note focuses primarily on three main areas of collaboration: (1) supporting the government to scope available payment mechanisms and develop a payment strategy, (2) training beneficiaries on the usage of the payment mechanism, and (3) preparing for scale-up. The field note ends by presenting some lessons learned that may be useful for stakeholders in other contexts planning similar initiatives.

STEP 1: SCOPING OUT POTENTIAL PAYMENT CHANNELS AND ANALYZING BENEFICIARY NEEDS

I. Scoping the market for payment service providers

When the World Bank and the IFC teams started thinking about payment channels for this program, the first step taken was to share with the Government an overview of current trends in safety net payments from countries such as Uganda, Haiti and Colombia. The purpose of this contextual knowledge exchange was to identify what experiences may be applicable in the Ivorian context to help start the program.

Markets like Côte d’Ivoire offer fragmented payment services. To identify suitable payment options for the national safety net program, a scoping study was done to better understand existing payment channels. It included not only a survey of payment service providers (e.g. banks, remittance companies, mobile network operators, the postal service, microfinance institutions, mobile money services) but also an extensive review of social protection programs in Côte d’Ivoire and internationally, including the legal framework surrounding them. IFC together with the World Bank presented these finding to the Government through a series of workshops that included detailed orientation on the mobile money market in Côte d’Ivoire.

Source: Global Findex 2017

50%

40%

30%

20%

10%

0%

Côte d’Ivoire

2014

4138

43

35

16

4245

2017 2014 2017 2014 2017 2014 2017 2014 2017 2014 2017 2014 2017

Benin BurkinaFaso

Mali Niger Senegal Togo

Financial institution account only Both FI & Mobile money account Mobile money account only

Mobile money accounts drove increases in account ownership in BCEAO overallAdults with an account (%)

4

The mobile-money market is relatively well developed in Côte d’Ivoire compared to other payment services assessed, including traditional banking. The relative strength of mobile money is supported by the good mobile network coverage, which serves 77 per cent of the population (with at least 2G internet) as of 201711. This high coverage, combined with favorable regulations, has enabled the mobile money services of mobile network operators (MMOs) to penetrate rural areas. The 2017 Global Findex reports a 38 percent mobile money account penetration in Côte d’Ivoire, with 31 percent penetration in rural areas alone, which is indicative of the very high uptake and relevance of mobile money among the target population of the social safety net program12. The regulatory environment also favors mobile money, with MNOs experiencing relatively few barriers to competition, while not sanctioning other types of digital financial services (DFS) that compete with mobile money in other markets.

II. Assessing the needs of beneficiaries of the social protection program

The challenge for the program was to find a solution that has a relatively soft learning curve and financially inclusive within the constraints that beneficiaries may face, such as lack of ID, limited literacy and numeracy, limited exposure to technology, etc.

To effect payment, a conservative choice such as a traditional, cash-in-hand payment mechanism might be the surest way to ensure that beneficiaries received their transfer reliably. However, payments in cash are more costly to the program compared to digital payments, and deprive beneficiaries of the means to use a transaction account, which has the advantage of offering secure storage, long-term savings, P2P transfers and other payments – while also providing a first step to more robust inclusion in the financial system as the market develops (loans, insurance, etc.)

STEP 2: SELECTING A PAYMENT SERVICES PROVIDER

A number of criteria were considered in selecting a payment services provider (PSP), including:

• Cost of the transfer • Transparency and the ability to monitor

payments effectively • Accessibility in rural areas • Security for beneficiaries • Ease of use for beneficiaries • The development impact of the

mechanism, including its potential to lead to financial inclusion

• Accessibility to those lacking ID documents

• Sustainability considerations, such as long-term profitability of G2P payments for the service provider and a pro-poor growth strategy

The Ministry of Employment and Social Protection in Cote d’Ivoire, which oversees this safety net program, opted to use a mobile money-based payment mechanism to deliver transfers for all program beneficiaries. Mobile money had a number of advantages over the other types of PSPs on the market. Unlike prepaid ATM cards and traditional cash payments, mobile money allows deposits to be made onto a full-featured transaction account. The transaction account was considered important to the program’s objective of exposing previously unbanked beneficiaries to financial services and leading to their inclusion in the formal financial sector over time.

While banks also offer payment onto a transaction account, they were deemed unsuitable for the program population due not only to the sparse coverage of branches and ATMs, but also the stringent Know Your Customer (KYC) requirement for formal identification that many program beneficiaries lacked. While microfinance institutions were another option considered by the program, there is only one MFI in Côte d’Ivoire with sufficient geographical coverage (UNACOOPEC), and this institution was experiencing financial difficulties. Mobile money thus became the only financial service offering true traceability, rigorous payment monitoring, and a distribution network across the country.

Among beneficiary individuals, average household size is large at 6.5, larger than the average for all households in the registry at 4.5 (containing poor and non-poor households). 48 percent are aged 15 or older, and 37 percent are literate. 71 percent of all heads of households hold an official government identification (National Identification Card and/or National Document issued by the National Identification Office).

11 World Bank analysis based on data from the Autorité de Régulation des Télécommunications/TIC de Côte d’Ivoire (ARTCI)12 World Bank (2017) Global Findex.

5

Figure 1: Digital transfers with Orange – How do payments happen?

11 World Bank analysis based on data from the Autorité de Régulation des Télécommunications/TIC de Côte d’Ivoire (ARTCI)12 World Bank (2017) Global Findex.

Account of the Ministry of Social Protection

Bank Bank

Account of Orange

01 02

E-money conversion

E-money Funds Pool

Orange money mobile money transfer system

Wire Transfer

Cash E-Money

Cash-Out

Orange money account

0304

06

e-wallet admin interface

Orange money agent

P1

P2

P3

04 05

There were initial concerns that not all beneficiaries would be reachable by MMOs. In particular, beneficiaries who live in isolated rural areas, who lack state-issued ID documents, and who cannot read and write, may have difficulty using a mobile-phone-based payment interface. However, it was ultimately decided that even such beneficiaries could be reached by MMOs if appropriate mitigation strategies were adopted. The various mitigation strategies adopted by the safety net program are discussed below.

The government ultimately opted to use Orange Money to manage all payments in the program. Orange was the clear market leader. While Moov and MTN both have a national presence, neither had a clear advantage over Orange in terms of either GSM network coverage or presence of mobile money agents in any of the program regions at the time of program start.

In addition, there were some program regions where Orange was the only mobile money provider with agent presence. This coverage advantage in rural areas was essential, since the program specifically targets the rural poor.

A multi-provider approach was also considered in order to give beneficiaries choice in providers and to accommodate beneficiaries that live in areas where the other two mobile money operators might have better coverage. However, it was decided that, until the safety net program could improve its capacity to manage payments – in particular, through the development of a dedicated computer module to manage payment lists – the advantages of using multiple PSPs were outweighed by the additional complexity involved.

6

STEP 3: DESIGNING CUSTOMIZED TRAINING FOR BENEFICIARIES

Using mobile phones to make social payments has a strong potential to increase financial inclusion in Côte d’Ivoire13. Many targeted program beneficiaries were already mobile phone users, while the rest were sensitized to begin using a basic feature phone given for free by the project (the cost to the project was approximately $10 per phone). Mobile money payments use ordinary mobile phones and do not require an Internet connection. The transactional account allows payments, transfers, remittances, savings, and the market in Côte d’Ivoire is quickly evolving to the point where microloans and other value-added services (VAS) may become more common place.

However, significant barriers remain to beneficiaries autonomously using mobile money. One such constraint is the beneficiaries’ ability to manipulate the mobile phone interface. Low literacy levels, poor and uncorrected eyesight, and a lack of experience with phone menu systems and USSD syntax impair usage. Poor numeracy can make it difficult to make financial decisions, for example how much to withdraw now and how much to save until later. Additionally, lack of understanding of the precise role of the PIN number in assuring the security of an account can lead beneficiaries to share PIN codes with others, compromising the accounts.

The only way to alleviate these constraints is to ensure that beneficiaries are adequately trained. Initially, many beneficiaries received their mobile phone on the day of the payment during the program launch ceremony, which did not give them time to familiarize themselves with the cash-out process nor the phone itself. To speed up payments, cash-outs were in fact often made by Orange staff on behalf of beneficiaries. This set-up made it possible for almost all participants to receive their entitlement on the launch day, but did not create lasting results for the program in terms of knowledge transfer.

To develop training materials that were adapted to the specific needs of the beneficiary population, the World Bank and IFC collaborated to study beneficiaries’ learning needs and to produce targeted training materials. The program leadership together with the World Bank/IFC team reviewed progress after the first payment was made by conducting both qualitative and quantitative research. They then corrected any issues with a particular emphasis on training. Three broad areas were identified for additional emphasis:

I. Limited literacy, numeracy and phone usage among participants.

The safety net program purposely targets poor and vulnerable families. This population group often has limited education and exposure to writing, including numbers. Most beneficiaries did not know when a message from the program had arrived, let alone how to locate it in the phone and read its content. Another common difficulty was understanding the sequence of digits to key into the phone to cash out. This was problematic for many because they simply could not recognize these digits on the handset. Many participants used their own ways of identifying numbers using their own imagination (linking numbers to agricultural tools or zero to the moon for instance). While some participants owned a phone prior to the program, they made a limited use of it. For new users, handling the handset was not intuitive. For example, some beneficiaries tended to use too much force when pressing the keys thus jamming them. As a result, a number of participants tucked away their phone after cashing out because they were fearful of mishandling it and thus block the SIM card.

Jeannette (53), Kanguierenou

“I would like to put my children’s numbers in my phone so I can call them, but I’m worried I might block the SIM card. I don’t use the phone, I go and get it charged in Brodo [the closest town]. Sometimes, I go myself, sometimes, my brother goes.”

II. Independent use of financial services may not be an end-goal desired by all or most safety net program participants.

Most participants welcomed the idea of learning a new skill, but others felt indifferent to this; either for lack of interest or due to a perceived inability to learn. Instead, some beneficiaries anointed a trusted relative who could handle the phone on their behalf while they themselves continued with their daily activities. Even among the participants who had used Orange Money for personal transfers prior to the program, there was a tendency to hand an agent the phone to carry out the required transactions to cash out. In the past, some of these participants would ask a trusted family member or friend to help execute the transfer. This meant that the opportunity afforded to participants to access a formal financial service independently was not as valued as expected.

13 There is much active investment in Côte d’Ivoire’s mobile infrastructure, including, for example, a World Bank-financed project aiming to bring coverage to the entire north of the country. See Lonie, Susie, Meritxell Martinez, Rita Oulai and Christopher Tullis (undated) Opportunities for Digital Financial Services in the Cocoa Value Chain. Insight from New Data. Côte d’Ivoire.

7

There could be many reasons for this, such as the fact that smallholder farmers may view the household as a unit and divide up tasks between people in the household to best manage everything that needs to be done – but still valuing access to the service, collectively. But independent use of a financial service is not prioritized.

One key take-away is the importance of taking into account this heterogeneity in beneficiaries’ preferences and abilities when developing training materials for a social protection project, or indeed its financial inclusion strategy. While some beneficiaries may be poised to benefit from a transactional account that can lead to financial inclusion, others may experience this as an unnecessary barrier to receiving their cash payment. Likewise, among beneficiaries interested in using financial services, not all are able to benefit from services of the same level of sophistication.

Chantal (52), Pliké-Somolo

“There is no network coverage. So I gave all my documents to the children of another beneficiary when they all went to the next village to cash out. It’s too far for me to walk to.”

III. The training methodology employed relied on visual aid on how to use a handset.

From the inception, there was little need to explain to participants the benefits of using mobile money and Orange. Their level of awareness and trust in the service was already very high. Many participants had good personal experience or positive feedback from their friends and relatives who used the service. Everyone could relate to aspects of convenience and security. What is more, almost all of them felt they could easily memorize a PIN to secure their electronic wallet.

Firmin (41), Kanguierenou

“Orange Money is about cash-in and cash-out. You can also send money to someone. At the bank, there is too much paperwork to fill out. You need to have a check to get money. With Orange Money there is less paperwork.”

Christine (42), Ankanzakro

“I trust Orange Money. My money is in my phone. You can’t take my money if you don’t know my secret code [PIN].”

In addition to the program presentations done by social workers, initial training material consisted of printouts of visuals of a handset, designed as a self-study tool to help participants familiarize themselves with the phone keypad and the digits. Subsequently, more emphasis was placed on community-based training and peer-to-peer learning since some beneficiaries – often older and uneducated – are used to turn to other community or family members for support. This confirms the positive role played by “helpers”.

Social workers have been equipped with large visual aids akin to a flipchart-size comic book, explaining the program objectives, offering basic numeracy tips, and detailing the cash out process (see annex for example training materials). For example, one visual depicts a program participant standing with an agent and keying into their phone the cash-out sequence in exchange for cash. The visual aids are complemented by audio recordings of the training script to ensure consistent messaging across all locations even after translation into various local languages. Since there is a strong need for on-going training support, existing village committees have also been trained on an abridged version of this material so that they can conduct refresher training.

STEP 4: PREPARING FOR SCALE-UP

The review of the pilot and the design of the customized comprehensive training was an opportunity to build a deeper understanding of the user experience and to try and address barriers to future sustained use of mobile money as a financial service, given the state of the ecosystem and regulatory concerns. Five key insights relevant to the scale-up phase were uncovered. Interestingly, some of these challenge established global best practice.

I. Number of payment service providers

When developing a payment strategy, social protection programs have to consider whether one PSP can reliably deliver payments to all program beneficiaries or whether multiple PSPs are needed. In some contexts, it may be that no single PSP has the capacity to manage all of the program’s payments, especially if the provider lacks a presence in certain program regions or if their staffing or liquidity are insufficient. However, even if a PSP with sufficient capacity to manage all payments can be found, it still may be desirable to contract with multiple PSPs. When multiple PSPs are involved, competition increases between providers, potentially decreasing costs to the program. This also helps the program respond

8

to beneficiaries’ individual preference by giving them a choice of providers. Finally, this can also help compensate for differential geographic coverage between PSPs, by, for example, allowing each beneficiary to choose the PSP who has an agent or a cash point nearest to their home.

Despite all the advantages of using multiple providers, the Ivoirian experience shows that it may still make sense for social protection programs to contract a single PSP. In Côte d’Ivoire, the social protection program considered contracting multiple PSPs but ultimately decided against it, both during the pilot phase of 5,000 beneficiary households and for the expansion to 35,000 households. There were two principal reasons why this decision was made:

• Limited management capacity. New social protection programs may not have the management capacity needed to manage payments made through multiple providers directly, and not all markets have a developed system of payments aggregators to provide this service. In particular, unless the program is using a robust digital management information system (MIS), the additional complication of managing multiple PSPs – such as generating multiple payment lists and monitoring data from multiple sources – may lead to increased errors and have negative impact on service delivery.

• Simplified payment experience for beneficiaries. From a training perspective, only one training script needed to be developed based on the cashing-out and other transactional processes of Orange. This also decreases confusion on the part of beneficiaries as peer support from other program beneficiaries is one of the key ways to get them in the habit of using the USSD syntax and other procedures necessary to access payments. In Côte d’Ivoire, as in many markets, each provider has a different USSD syntax, which is already confusing. In terms of customer service, having one call center and one support team that handled all enquiries streamlined communication.

II. Regulation

Some regional disparities were observed among the beneficiary population in terms of access to a valid government-issued ID. Although nearly all beneficiaries understood the need to have an official ID in their daily lives and were willing to set aside part of their transfer to obtain one, they did not all possess the right documentation such as a birth certificate to obtain a legal ID.

Since this is an issue in many African countries and not solely in Côte d’Ivoire, serving the poorest populations remains a challenge when program participation is conditional on having an ID. The imperative of respecting KYC requirements and anti-money laundering and counter-financing of terrorism (AML/CFT) rules must be balanced against the need to foster the financial inclusion of the underserved.

The Ministry of Employment and Social Protection, which champions the safety net program, had a multi-pronged approach for the pilot program that allowed 100 percent of beneficiaries to receive their transfer using mobile money. The program design was adapted in specific ways to accommodate the fact that not all potential recipients possessed the required documents to fulfill KYC requirements. Crucially, the program allowed beneficiary households to designate trusted non-household members to receive the program benefit onto their mobile money account (see Box 1 for details. )

9

BOX 1. STRATEGY IMPLEMENTED BY THE PRODUCTIVE SOCIAL SAFETY NET PROJECT (PSNP) FOR EXPANDING ACCESS TO MOBILE PAYMENTS TO HOUSEHOLDS WITHOUT ID Since not all program beneficiaries had access to ID documents acceptable to the MMO for KYC purposes, it was necessary to make some modifications to the design of the PSNP in order to allow all beneficiaries to be paid through the mobile channel. These modifications and exceptions were:

1. Temporary KYC exception. Initially, the program negotiated a temporary exemption from the MMO’s usual KYC requirements to allow beneficiaries to open accounts using the ID card issued by the program instead of the usual national IDs.

2. Support for getting an ID. Beneficiaries were informed of the need to acquire a state-issued ID to continue receiving their benefits, and were encouraged to use their initial benefits to finance the cost of doing so.

3. Modifying program design to accommodate those without access to ID. Since not all beneficiaries have access to the documentation needed to acquire an ID, even once they understand the benefits, the PSNP allowed households flexibility in designating a trusted transfer recipient.

• Male recipients. Although PSNP households are normally encouraged to nominate a female recipient, the PSNP elected to allow the household to designate a male recipient on an exceptional basis if no women in the household had an official ID.

• Community recipients. In cases where no one in the household had an official ID, the PSNP allowed the household to nominate a trusted person in the community (outside of the household) on an exceptional basis. Such households are flagged for additional monitoring to ensure that beneficiaries really receive the transfer. Data suggests that initial concerns of fraud stemming from this exception may be overblown.

4. Collaborating with ongoing reform of the national ID system. Côte d’Ivoire is currently in the process of reforming its national ID system to expand access, including in rural areas and to non-citizens, and to offer low-cost and digital credentials that can be used to access public services. The PSNP is coordinating closely with this effort to ensure that its beneficiaries are among the first to receive these new credentials.

As a result of this strategy, the PSNP was able to pay 100 percent of its beneficiaries onto a mobile money account in early 2018, thereby overcoming KYC as a binding constraint to digital social payments in the short term and laying the groundwork for full financial inclusion of project beneficiaries in the long term.

In the long term, it is desirable for the program to have the latitude to choose transfer recipients based on factors linked to development outcomes and not have recipient choice be motivated by a need to mitigate against regulatory barriers.14 While the appropriate long-term solution remains a combination of risk-based KYC regulation15 combined with increasing access to official IDs, the innovative approach taken by the PSNP shows a strategy for how digital payments can be implemented in social payments programs in contexts where these barriers have yet to be overcome.

14 Check systematic review exploring how women recipients yield better development outcomes when receiving conditional cash transfers. https://assets.publishing.service.gov.uk/media/57a08a6aed915d622c00070d/EconomicTtransfer2012Yoong.pdf Also how the bargaining position of women improved after they received transfers, and that women were then better able to influence the pattern of expenditures at https://openknowledge.worldbank.org/bitstream/handle/10986/7471/WPS4282.pdf?sequence=1

15 A recent study on the regulatory environment for DFS in Côte d’Ivoire shows that while risk-based KYC is already adopted in regulation from the BCEAO, that this regulation conflicts with and is nullified by conflicting regulatory requirements. http://www.cgap.org/sites/default/files/Working-Paper-Market-Stystem-Assessment-of-Digital-Financial-Services-in-WAEMU_0.pdf

10

III. Agent network

For beneficiaries to be able to realize the benefits of the social transfer, they need to be able to turn it into cash. The relatively weak presence of mobile money agent networks in some program areas, as well as the liquidity constraints of some agents, are challenges. Even in countries like Côte d’Ivoire, with relatively high rural mobile money penetrations (94,443 mobile money agents as of December 2017, or approximately 1 agent for every 89 active Ivoirians mobile money users)16 there is still low agent coverage in the poorest regions.

To supplement these weak fixed agent networks, the program elected to have the PSP deploy roaming agents that would travel to the villages during the periods in which the program makes its transfers. This method allowed the program to ensure that all beneficiaries were able to access their benefits in cash in a timely fashion and without excessive travel costs or waiting times. However, there are also disadvantages to such an approach. Deploying itinerant agents can be costly compared to using existing agent infrastructure. Additionally, the fact that this cash-out capacity is only available during specific periods means that beneficiaries cannot cash out on demand, which may limit the utility of their accounts for managing day-to-day finances. If cash-outs are only possible on fixed “payments days” attended in person by program staff, then the mobile money platform risks becoming simply an alternate cash-delivery mechanism with few benefits beyond traditional cash payments.

This means that robust and accessible agent networks are needed if beneficiaries are to fully benefit from having a financial account. In the short term, stopgap measures such as the deployment of itinerant agents may be necessary. In the long term, however, social protection programs should conceive of themselves as key stakeholders in the mobile money sector and should actively attempt to encourage providers to expand agent networks.

These programs generally generate large, predictable cash flows in well-defined geographical regions. Such cash flows can help improve the business case for deploying agents to program regions. Social protection programs should work with PSPs to understand their needs and business models, and actively communicate a program’s medium- and long-term strategies to PSPs to allow them to integrate these into their own expansion strategies. Social protection programs should also look for opportunities to help incentivize MMOs to invest in agent networks. For example, it might be possible for the program to cluster beneficiaries who receive payments through a given PSP geographically in order to improve the business case to that PSP to deploy an agent in that area. When liquidity management is identified as a constraint, the program could also consider spreading payments out over time so that not all beneficiaries in a given community receive payments on the same day. Such a practice could help local agents cope more effectively with program cash flows, improving overall liquidity while also providing more steady business throughout the month. In general, it may help to view investments made by MMOs in rural distribution networks also as investments in the general financial infrastructure of these communities.

Social protection programs should recognize that their programs account for a significant share of the digital cash flow in many areas and work actively to use this market power to incentivize investments in rural financial infrastructure that will pay significant dividends not only for the delivery of the social protection program itself but also the broader financial lives of program beneficiaries.

16 BCEAO (2018) Etat Des Services Financiers Par Téléphonie Mobile dans L’UEMOA en 2017

11

CONCLUSION

Currently, the productive social safety net program is unconditional in nature, with households receiving quarterly transfers of FCFA 36,000 (approximately US$ 72). This transfer amount is the equivalent of 15 percent of consumption of average poor households, and 21 percent of consumption of extremely poor households. The program covers 35,000 households as of June 2018 (194,000 individuals, 2.0 percent of the poor) from a social registry managed by the program containing 92,000 households across 12 regions in the Northern, Western and Central zones and has been extended to cover 80,000 additional beneficiaries by 2022.

There is much debate about the sustainability of social protection programs targeting the poor, and on the lasting impact on beneficiaries once they stop receiving program benefits. To address these sustainability concerns, many social protection programs have begun to encourage beneficiaries to use their program benefits to invest in productive activities that can continue to bear fruit after the program ends. In the same vein, if social protection programs could have the effect of bringing previously unbanked populations permanently into the financial system, then the continued benefits of financial inclusion could be one of the legacies of the program.

Merely paying benefits onto a digital account is not sufficient to lead to lasting financial inclusion. International experience shows that additional support activities need to be added. These may include, for example, strategic support to MMOs, through public expenditure to improve financial infrastructure, such as the expansion of the agent network, as mentioned above. Beneficiaries also need training to understand how to use financial services. Even if beneficiaries are not active users of DFS at the outset of a social protection program, the program offers a prolonged, often multi-year engagement with beneficiaries that sees them regularly using a digital account to receive payments.

Social protection programs should take advantage of this sustained engagement to train beneficiaries gradually over time, building their capacity to use DFS autonomously. This way, by the time that they exit the program, those beneficiaries who desire to and can benefit from using DFS to improve their financial lives – by saving, budgeting, transferring money, or even applying for digital microloans as the market matures – can do so.

The best suited approach is to have program staff supervise payments in the early days and phase out this support over time to avoid any risk of fraud, to keep cost down, and to foster autonomous use and financial inclusion. Though initially labor intensive, this kind of approach may well be the best model to promote sustained adoption of digital financial services by the poorest.

RECOMMENDATIONS

At a minimum, in Côte d’Ivoire, we recommend that as it scales up, the safety net program work to ensure that:

1) Cash-out points are accessible and located near beneficiaries’ homes;

2) Beneficiaries can use the product autonomously and know where to go for help when needed;

3) Mobile money agents have a business case for ongoing presence in beneficiary localities to ensure the sustainability of cash-out points;

4) MMOs have a business case for ongoing relationship to the program (fees at a level that allows for investment, opportunity to cross-sell value-added services, visibility concerning the program’s long-term plans, etc.);

5) Beneficiaries are supported with obtaining an ID, needed to meet KYC requirements, over the long-term;

6) A multi-payment provider approach is introduced once existing data management capacity and MIS deficiencies are solved.

ANNEX 1: PICTURES OF THE BENEFICIARY TRAINING MANUAL

Presentation of the program

Advice on how to secure the PIN

Introduction to the phone touchpad

Contact the Publisher:

Anna Koblanck

AKoblanck@ifc.org

+27(0) 11-731-3000

IFC, Sub-Saharan Africa

14 Fricker Road, Illovo,

Johannesburg

AUTHORS

Christopher Tullis is a Social Protection Consultant working with the World Bank.

Heba Elgazzar is a Senior Economist working for the World Bank.

Kokoevi Sossouvi is a Digital Financial Services Consultant working with the World Bank and IFC.

Meritxell Martinez is a Project Lead working for IFC Advisory Services on Digital Financial Services in social services and agri value chains.

AcknowledgementsThe team would like to acknowledge the following people for the guidance and support during the implementation of the program and the development of this concept note: Riadh Naouar, Anna Koblanck, Rita Oulai, and Lesley Denyes from IFC, and the partners from the Ministry of Employment and Social Protection in Cote d’Ivoire.

DECEMBER 2018